Key papers, reviewed

Defined the cholinergic anti-inflammatory pathway: vagal efferents → splenic nerve → α7-nAChR on macrophages → suppressed TNF release. Reframed the vagus as an immunological circuit, not just an autonomic one.

• 01Vagus nerve actively restrains innate immune cytokine output
• 02α7 nicotinic receptor identified as the molecular target on macrophages
• 03Provided the rationale for VNS in RA, sepsis, IBD

Foundational but mechanistic, not clinical. Translation to chronic human disease has been slower and noisier than the elegant mouse model implies.

Two double-blind RCTs (E03, E05) in drug-resistant focal epilepsy comparing high vs low VNS parameters. Provided the basis for FDA approval of LivaNova VNS in 1997.

• 01~25–30% mean seizure reduction at 3 months in high-stimulation arm
• 02Roughly one third of patients achieved ≥50% reduction (responders)
• 03Voice alteration, cough, dyspnea most common adverse effects

Active-control design (not true sham); chronic open-label extensions inflated apparent benefit due to titration and placebo effects.

10-week sham-controlled trial of adjunctive VNS in TRD. Failed its primary acute endpoint; long-term open-label data later supported the 2005 FDA approval — a controversial decision.

• 01No statistically significant acute response over sham
• 0212-month open-label cohort showed continued improvement
• 03Drove FDA approval based on durability rather than acute efficacy

Approval based on uncontrolled long-term data is methodologically weak. CMS later restricted reimbursement, and clinical use stayed niche until RECOVER.

Largest sham-controlled VNS-in-TRD trial to date (~500 patients). Designed to settle the long-running question after the equivocal 2005 data.

• 01Modest but statistically significant improvement on multiple secondary outcomes at 12 months
• 02Primary acute endpoint again did not separate cleanly from sham
• 03Helped restore CMS coverage decision for VNS in TRD

Effect sizes remain modest; surgical/cost burden is high. Best framed as 'last-resort augmentation' not first-line.

First-in-human chronic VNS in RA patients. Demonstrated reduced TNF and clinical improvement (DAS28) in a small cohort, validating Tracey's mechanism in humans.

• 01Significant reduction in LPS-induced TNF in whole-blood assay
• 02Clinical DAS28 improvements in majority of patients
• 03Effects reversible with stimulator OFF

Small (n≈17), open-label, no sham. Hypothesis-generating; not actionable evidence on its own. Set the stage for SetPoint's RESET-RA.

Multicenter sham-controlled RCT pairing implanted VNS with task-specific upper-limb rehab in chronic stroke. Led to FDA approval (2021) of MicroTransponder Vivistim.

• 01VNS-paired rehab roughly doubled clinically meaningful FMA-UE response rate
• 02Effect persisted at 90-day follow-up
• 03Mechanism likely cortical plasticity gating via locus coeruleus / basal forebrain

Effect impressive but in a selected chronic-stroke population; generalizability and cost-effectiveness debates ongoing.

Two sham-controlled RCTs of nVNS (gammaCore) in cluster headache. Episodic cluster benefited; chronic cluster did not separate from sham.

• 01Higher pain-free rate at 15 min in episodic cluster vs sham
• 02Drove FDA clearance of gammaCore for cluster
• 03Chronic-cluster failure tempers enthusiasm

Sham device design hard to perfect (users can perceive stimulation). Real-world use limited by cost and prescription access.

Systematically tested ear-electrode positions during fMRI to identify which loci most strongly activate brainstem vagal nuclei (NTS, locus coeruleus). Cymba conchae outperformed tragus and earlobe.

• 01Cymba conchae produces strongest classical vagal central activation
• 02Earlobe (common sham) does not activate NTS — useful control site
• 03Provides anatomical rationale for taVNS electrode placement standards

Small n; fMRI is an indirect proxy. Doesn't prove clinical benefit, only neural engagement.

Aggregated taVNS HRV studies. Found inconsistent and small effects across protocols, parameters, and populations.

• 01Acute HRV effects are heterogeneous and often null
• 02Parameter choice (frequency, intensity, electrode site) drives variance more than disease state
• 03Suggests current 'taVNS = vagal tone boost' marketing oversells the data

High between-study heterogeneity limits pooling. Useful as a sober counterweight to wellness claims.

Articulates polyvagal theory: a 'social engagement system' built on a phylogenetically newer myelinated ventral vagal complex distinct from older unmyelinated dorsal vagus.

• 01Proposes hierarchical autonomic states: ventral vagal (safe), sympathetic (mobilized), dorsal vagal (shutdown)
• 02Hugely influential in trauma therapy, somatic practices, and wellness culture

Comparative-anatomy claims (e.g., reptiles lack myelinated cardiac vagus) have been challenged by Grossman, Taylor and others. Treat as a useful clinical heuristic, not validated neuroscience.

Systematic critique arguing that all five core premises of polyvagal theory are either unsupported or contradicted by current comparative cardiovascular neuroscience.

• 01Myelinated cardiac vagal fibers are not unique to mammals
• 02RSA is not a clean index of 'ventral vagal' tone or social engagement
• 03Clinical popularity of polyvagal language outpaces its scientific validity

Authoritative scientifically; clinicians using polyvagal language should be aware that it is a metaphor, not a settled theory.

Daily 1-hour tragus taVNS over 6 months reduced AF burden and inflammatory cytokines in paroxysmal AF patients.

• 0185% reduction in AF burden vs sham at 6 months
• 02Reduced TNF-α and CRP
• 03Supports inflammatory-reflex involvement in AF substrate

Single center, modest n, surrogate endpoint (AF burden on monitor). Larger pivotal trials needed before guideline change.

Small sham-controlled trials testing taVNS for fatigue, cognitive symptoms, and dysautonomia in post-acute COVID-19.

• 01Signals of benefit on fatigue and mental clarity
• 02Tolerability good; dropout low
• 03Supports rationale for ongoing larger trials (Parasym, NIH RECOVER-NEURO)

Small samples, subjective outcomes, high placebo expectancy in long-COVID populations. Promising but very preliminary.

AAN systematic review and guideline for VNS in epilepsy: response rates increase over time, may improve mood, and is generally safe long-term.

• 01Level B evidence for improved seizure outcomes over months–years
• 02Level B for mood improvement independent of seizure change
• 03Level C for use in children and Lennox-Gastaut syndrome

Pre-RECOVER, pre-closed-loop AspireSR. Useful baseline; clinicians should pair with newer registry data.